Public Lab Research note


Uncontrolled reentry

by cfastie | October 25, 2015 05:16 25 Oct 05:16 | #12333 | #12333

Above: Stereographic projection of a panorama stitched from 22 photos taken by a Canon S100 on a Saturn V autoKAP rig lofted by a Fled kite on October 23, 2015.

The sun finally came out at 4:00 PM on Friday, so I ran up the hill and had the Fled in the air at 5:00. The temperature was 40°F and falling, so it was a good chance to test if rechargeable eneloop batteries could power the SkyShield when it was chilly. There were six of them in the new two-case battery pack.

PetriTub-679-2.jpg
Above: The sun was setting fast and in the rush to launch I forgot to check the protective tubing which cheerfully reveals itself in every photo taken at the uppermost tilt position.

The winds were swirlier than any I have flown a kite in because they were coming right over the 500 foot tall hill next to the clearing. I could not get the Levitation delta launched, but the Fled went up quickly and I attached the Saturn V Rig as soon as I could. Five hundred feet up the winds were plenty strong for the Fled, and they did not wane as the sun set.

PetriTub-679-816.jpg
Above: This is the first time I have captured a sunset from a kite lofted camera. This photo was very underexposed and was lightened in Photoshop.

PetriTub440-450cu2k.jpg
Above: Before the sun set several hundred photos of late fall foliage were captured as the Fled flew downwind of the clearing. Stitched from five photos. The color was not altered before or after stitching.

When I started to haul in the Fled after 45 minutes, the wind was fierce and the Fled was pulling hard. I had been struggling to haul in line for 10 minutes when suddenly it got a lot easier to wrap the line around the spool. When the Fled dropped into the lee of the hill, it got hit by an eddy from another direction and started to fall. I have never seen so much kite line just fall to the ground in front of me. Wrapping it around the spool or running had no effect on the kite or the plummeting camera. The Saturn V Rig hit the ground pretty hard.

GPSFallgraph.jpg
Above: The SkyShield was triggering the shutter every three seconds, and the camera was recording the GPS coordinates and altitude for each photo. According to the GPS data, during the final 40 seconds of the descent, the camera fell two feet per second, twice as fast as during the previous minute and a half.

LastThreePhotos.jpg
Above: The last two photos before the rig hit the ground (top) and the first photo taken on the ground (bottom). The GPS altitude data may not be very accurate because the camera was falling fast. These photos were taken about three seconds apart.

Before eight photos had been taken by the grounded rig (i.e., 24 seconds after the crash) I had the SkyShield and camera shut down and was trying to keep the Fled from falling into the forest. Fortunately the eddy was blowing the kite back "upwind" into the clearing. So by sheer good luck the Fled landed safely. The camera also landed safely despite the crash. The shock was absorbed when the plastic frame of the Saturn V Rig broke cleanly at the leg attachment.

SatVerg-681-20.jpg
Above: The oak legs hit the ground first and the plastic frame snapped where the leg bracket bolts onto it. The camera and camera tray probably never touched the ground.

SatVerg-681-27.jpg
It took about 10 minutes to repair the frame with a soldering iron and some PLA filament. It's probably stronger now than it was new.

So the Saturn V Rig performed just as it was designed to. Cris Benton refers to the legs on his KAP rigs as "sacrificial." A primary role of the entire rig is to protect the camera, so I guess the whole thing is sort of sacrificial. The S100 camera did not suffer at all and took 1300 photos during a flight of the repaired rig the next day.

Above: Half-spherical panorama stitched from 22 photos taken during this flight.

Anyway, the goal of this exercise was supposed to be testing eneloop batteries. After the 50 minute flight, I walked home and turned the SkyShield on again. It was 35°F and I hung it outside while I ate dinner. It ran fine for another hour. So I think six rechargeable AAA batteries are going to be my default power supply for the SkyShield until it warms up again next summer.

Flight notes:

Camera:

  • Model: PowerShot S100
  • ISO: 200
  • Shutter speed: 1/500 second (Tv)
  • Focus: manual on infinity
  • Focal length: 24mm (eq.)
  • White balance: Sunny
  • GPS: on

CHDK:

  • Remote enabled: one push, quick
  • The S100 has a shutter priority mode so CHDK is not needed for that

SkyShield:

  • Version: 2.4 (3-switch DIP)
  • Sketch: version 2.08sfm (beta)
  • Mode: Mode 0 (for 24 mm lens, 25 photos/cycle)
  • Power: 6 eneloop AAA
  • Customization: The nadir tilt angle was changed from 20 to 11 so the camera was pointed straight down.

Flight:

  • Kite: Fled
  • Wind: From N, 12 to 20 mph
  • Duration: 50 minutes
  • Photos taken: 1005

Panorama stitching:

  • Software: Microsoft ICE
  • Post processing: ICE's "auto complete" feature was used to create extra sky to fill in missing areas in the panoramas. Some stitching errors in the little planet panorama were hidden in Photoshop. Panoramas at Photosynth cannot be passed through Photoshop so what you see there is what Microsoft ICE produced.

3 Comments

Haha, great documentation of a rough flight! On the battery front, have you considered using hand warmers to keep the batteries warm? Oliver Yeh used to do that.

Is this a question? Click here to post it to the Questions page.

Reply to this comment...


That's a great idea. Last week as I was driving away from the house on the way to Middlebury College to launch the Saturn V Rig at 35°, It occurred to me that I should have grabbed a few foot warmers from my huge box of them. It's windy up there near the kite, so some insulation might be needed too. At this point I think adding a couple of batteries is the smartest solution. Six volts doesn't leave enough margin, but seven to nine volts seems to do the trick.

Reply to this comment...


Great report of the incident and crash performance of the rig, also beautiful imagery.

Reply to this comment...


Login to comment.